Photolysis of r-Azidoacetophenones: Direct Detection of Triplet Alkyl Nitrenes in Solution Pradeep N. D. Singh, Sarah M. Mandel, Rachel M. Robinson, Zhendong Zhu, Roberto Franz, Bruce S. Ault, and Anna D. Gudmundsdo´ttir* Department of Chemistry, University of Cincinnati, Cincinnati, Ohio 45221-0172
[email protected] Received May 20, 2003
We report the first detection of triplet alkyl nitrenes in fluid solution by laser flash photolysis of R-azido acetophenone derivatives, 1. Αzides 1 contain an intramolecular triplet sensitizer, which ensures formation of the triplet alkyl nitrene by bypassing the singlet nitrene intermediate. At room temperature, azides 1 cleave to form benzoyl and methyl azide radicals in competition with triplet energy transfer to form triplet alkyl nitrene. The major photoproduct 3 arises from interception of the triplet alkyl nitrene with benzoyl radicals. The triplet alkyl nitrene intermediates are also trapped with molecular oxygen to yield the corresponding 2-nitrophenylethanone. Laser flash photolysis of 1 reveals that the triplet alkyl nitrenes have absorption around 300 nm. The triplet alkyl nitrenes were further characterized by obtaining their UV and IR spectra in argon matrices. 13C and 15N isotope labeling studies allowed us to characterize the C-N stretch of the nitrene intermediate at 1201 cm-1. The pursuit of organic magnetic materials has sparked renewed interest in triplet nitrene intermediates, which are ideal candidates because of their high-spin properties.1 It is presumed that the direct photolysis of alkyl azides does not lead to the formation of triplet alkyl nitrene intermediates; instead, alkyl azides are considered to undergo a rearrangement from the excited state of the azide to form imine derivatives with loss of nitrogen.2 This is supported by the fact that there are only a few examples where products have been isolated that could be attributed to the decomposition of alkyl azides into alkyl nitrene intermediates.3 However, physical measurements have been reported that support the existence of triplet alkyl nitrenes in matrices and in the gas phase.4-8 Wasserman, for example, obtained EPR spectra of triplet alkyl nitrenes in a matrix by sensitized photolysis of alkyl azides.4 Engelking and co-workers (1) (a) Rajca, A. Chem. Rev. 1994, 94, 871. (b) Iwamura, H. Adv. Phys. Org. Chem. 1990, 26, 179. (c) Magnetic Properties of Organic Materials; Lahti, P. M., Ed.; Marcel Dekker: New York, 1999. (d) Nicolaides, A.; Tomioka, H.; Murata, S. J. Am. Chem. Soc. 1998, 120, 11530. (e) Subhan, W.; Rempala, P.; Sheridan, R. S. J. Am. Chem. Soc. 1998, 120, 11528. (2) (a) Abramovitch, R.; Kyba, E. P. J. Am. Chem. Soc. 1971, 93, 1537. (b) Moriarty, R. M.; Serridge, P. J. Am. Chem. Soc. 1971, 93, 1534. (c) Moriarty, R. M.; Reardon, R. C. Tetrahedron 1970, 26, 1379. (3) (a) Banks, R. E.; Berry, D.; McGlinchey, M. J.; Moore, G. J. J. Chem. Soc. C 1970, 1017. (b) Pancrazi, A.; Khuong-Huu, Q. Tetrahedron 1975, 31, 2041 (4) (a) Barash, L.; Wasserman, E.; Yager, W. A. J. Am. Chem. Soc. 1967, 89, 3931. (b) Wasserman, E.; Smolinsky, G.; Yager, W. A. J. Am. Chem. Soc. 1964, 86, 3166. (c) Smolinsky, G.; Wasserman, E.; Yager, W. A. J. Am. Chem. Soc. 1962, 84, 3220. (5) Radziszewski, J. G.; Downing, J. W.; Jawdosiuk, M.; Kovacic, P.; Michl, J. J. Am. Chem. Soc. 1985, 107, 594. (6) (a) Chapell, E. L.; Engelking, P. C. J. Chem. Phys. 1988, 89, 6007. (b) Carrick, P. G.; Brazier, C. R.; Bernath, P. F.; Engelking, P. C. J. Am. Chem. Soc. 1987, 109, 5100. (c) Carrick, P. G.; Engelking, P. C. J. Chem. Phys. 1984, 81, 1661.
allowed methyl azide to react with metastable N2 in a flowing afterglow and were then able to obtain and analyze the UV emission spectrum of triplet methyl nitrene in the gas phase.6 Ferrante8 and Platz et al. have obtained a UV absorption spectrum of triplet alkyl nitrene in matrices at low temperature. 9
By photolyzing azidoaryl ketone derivatives, which contain an intramolecular triplet sensitizer, we have succeeded in obtaining products that can be attributed to the trapping of alkyl nitrenes in bimolecular reactions.10 The intramolecular triplet sensitization ensures efficient formation of the triplet alkyl nitrene by bypass(7) (a) Shang, H. R.; Gao, R. M.; Ying, L. M.; Zhao, X. S.; Tang, Y. Q. Chem. Phys. Lett. 1997, 267, 345. (b) Ying, L. M.; Xia, Y.; Shang, H. R.; Zhao, X. S.; Tang, Y. Q. J. Chem. Phys. 1996, 105, 5798. (8) Ferrante, R. F. J. Chem. Phys. 1987, 86, 25. (9) Gritsan, N. P.; Likhotvorik, I.; Zhu, Z.; Platz, M. S. J. Phys. Chem. A 2001, 105, 3039. (10) Mandel, S. M.; Krause Bauer, J. A.; Gudmundsdottir, A. D. Org. Lett. 2001, 3, 523.
10.1021/jo034674e CCC: $25.00 © 2003 American Chemical Society
Published on Web 09/25/2003
J. Org. Chem. 2003, 68, 7951-7960
7951
Singh et al. SCHEME 1
SCHEME 2
ing the singlet manifold of the azide precursor. Triplet alkyl nitrenes are presumably the ground state, as triplet methyl nitrene is 31 kcal/mol lower in energy than singlet methyl nitrene.11 Since a spin barrier prohibits the triplet nitrenes from rearranging to singlet imine products, triplet alkyl nitrenes can be expected to be long-lived and thus directly detectable. In this present study we have extended our previous work10 and set out to detect triplet alkyl nitrenes directly in fluid solution with time-resolved laser flash photolysis and to further characterize these intermediates by obtaining their UV and IR spectra in matrices. We specifically wanted to characterize the C-N stretch in the alkyl nitrene intermediates by using 15N and 13C labeling.
(11) Travers, M. J.; Cowles, D. C.; Clifford, E. P.; Ellison, G. B. J. Chem. Phys. 1999, 111, 5349.
7952 J. Org. Chem., Vol. 68, No. 21, 2003
Results and Discussion Product Studies. We photolyzed azides 1a-f in toluene solutions at ambient temperature and kept the conversion below 20%. In all instances amide 3 was formed as the major product along with some acetophenone 4 and benzaldehyde, 5 (see Scheme 1). We have previously reported the product studies from photolysis of 1a and 1c but these compounds are included here for comparison.10 Formation of amide 3 suggests that azides 1 undergo R-cleavage to form benzoyl radicals in competition with nitrene formation (see Scheme 2). Benzoyl radicals do not readily abstract H atoms from appropriate solvents but are quenched rapidly by molecular oxygen.12 (12) (a) Neville, A. G.; Brown, C. E.; Rayner, D. M.; Lusztyk, J.; Ingold, K. U. J. Am. Chem. Soc. 1991, 113, 1869. (b) Brown, C. E.; Neville, A. G.; Rayner, D. M.; Ingold, K. U.; Lusztyk, J. Aust. J. Chem. 1995, 48, 363.
Photolysis of R-Azidoacetophenones SCHEME 3
SCHEME 4
SCHEME 5
Thus 3 can be formed similarly from interception of triplet nitrene, 2, by a benzoyl radical, whereas benzoyl radicals slowly abstract an H atom from toluene to form a minor amount of benzaldehyde, 5. It is also possible that triplet alkyl nitrenes 2 fragments into benzoyl and imine radical. Acetophenone 4 can be formed as a secondary photoproduct from photolysis of azo dimer 6, which is presumably formed by dimerization of two triplet nitrene intermediates. We were, however, not successful in isolating azo dimer 6, presumably because the concentration of 6 is minimal since it reacts efficiently. Nevertheless, triplet alkyl nitrenes can be expected to dimerize to form azo dimers in a manner similar to triplet phenyl nitrenes.13 It is, however, also possible that acetophenone is formed from β-cleavage of azide 1a. It is also likely that 3 can be formed through a nonnitrene mechanism if the benzoyl radical formed from the irradiation of azide 1 reacts with a second azide molecule (see Scheme 3). To test this hypothesis we formed benzoyl radicals in the presence of admantanyl (13) (a) Schrock, A. K.; Schuster, G. B. J. Am. Chem. Soc. 1984, 106, 5228. (b) Marcinek, A.; Levya, E.; Whitt, D.; Platz, M. S. J. Am. Chem. Soc. 1993, 115, 8609. (c) Schuster, G. B.; Platz, M. S. Adv. Photochem. 1992, 17, 70.
azide, 7. Since azide 7 does not absorb light above 350 nm, benzoyl radicals can be formed in its presence by irradiating through the 349 nm filter benzoin methyl ether, 8, that undergoes R-cleavage readily. We specifically selected 8 as the precursor for the benzoyl radical, since its excited state has been estimated to decay with a rate of more than 1010 s-1 and thus intermolecular triplet sensitization of azide 7 with 8 can be minimized.14,15 Photolyzing 8 in the presence of 7 did not lead to any azide reactivity, but only the expected photoproducts from 8 (see Scheme 4). The photoproducts from 8, 5, 9, 10, 11, and 12 were formed in the same ratio as when azide was not present. In contrast, we have shown that triplet-sensitized photolysis of azide 7 mainly leads to formation of di-adamantan-1-yl-1,2-diazene, presumably via triplet adamantyl nitrene formation.15 Effect of Temperature on Product Formation. We investigated how the temperature of the photolysis affected the product ratio (see Scheme 5). Photolysis at 100 °C yielded benzaldehyde and 3a as at room tem(14) Lewis, F. D.; Lauterbach, R. T.; Heine, H. G.; Hartmann, W.; Rudolph, H. J. Am. Chem. Soc. 1975, 97, 1519. (15) R. F. Klima and A. D. Gudmundsdottir. Unpublished results. Triplet sensitization of 1-azidoadamantane with acetophenone yields mainly di-adamantan-1-yl-1,2-diazene.
J. Org. Chem, Vol. 68, No. 21, 2003 7953
Singh et al. SCHEME 6
pereature, but in addition products 9, 10, and 12 were formed. These compounds must originate from benzoyl radicals. Dibenzyl 9 is presumably formed from dimerization of two benzoyl radicals, and 10 must come from trapping of a benzoyl radical with toluene radicals. Similarly, 13 must come from trapping of the nitrene 2a with two benzoyl radicals. Thus at higher temperature R-cleavage must be favored over triplet energy transfer since more of the products can be attributed to formation of benzoyl radicals. Photolysis at -63 °C gave 4a, 3a, and 12 as the major products. Diketone 12 must come from dimerization of two acetophenone radicals. Hence, photolysis at lower temperature yielded less of products coming from R-cleavage and more of products that must come from formation of acetophenone radicals. The product ratio indicates that the activation energy of R-cleavage must be higher than that for energy transfer to form triplet alkyl nitrene and the presumable β-cleavage. Oxygen Trapping of the Reactive Intermediates Formed upon Photolysis of Azide 1a. We studied the photolysis of azide 1a in oxygen-saturated solutions to trap the intermediates formed with oxygen. Irradiating azide 1a in an oxygen-saturated toluene solution at ∼25 °C yielded benzoic acid as the major product and lesser amounts of acetophenone 5a and 2-nitrophenylethanone, 15 (see Scheme 6). Formation of benzoic acid is expected since benzoyl radicals are efficiently trapped by molecular oxygen.12 Formation of 15 demonstrates that we have trapped triplet alkyl nitrene 2a with oxygen just as triplet aryl nitrenes are trapped with molecular oxygen to form nitroaryls.16 Photolysis at ca. -63 °C of azide 1a in toluene saturated with oxygen yielded nitro compound 15 as the major product. Thus, the major intermediate formed at ca. -63 °C must be nitrene 2a, which is trapped by molecular oxygen to yield 15. Additionally, since the combined yield of 4a and 12 decreases in the presence of molecular oxygen we can speculate that these acetophenone-based products are not formed directly by β-cleavage of 1 but rather result from nitrene 2a. Presumably, nitrene 2a undergoes dimerization to form 6 in the absence of molecular oxygen, whereas in the presence of oxygen nitrene 2a is trapped and thus any secondary photoproducts that might come for dimer 6 are decreased. Electron Configuration of the Excited States of Azide 1 and Product Formation. We analyzed the product formation taking into account the electron configuration of the triplet excited state of the ketone in azides 1, since R-cleavage is generally thought to arise from the n,π* excited state, whereas triplet energy transfer can take place from both the n,π* and the π,π* (16) Abramovitch, R. A.; Challand, S. R. J. Chem. Soc., Chem. Commun. 1972, 964.
7954 J. Org. Chem., Vol. 68, No. 21, 2003
manifolds. We estimated that the triplet energy and the electronic configuration of the lowest excited ketone in azides 1 are the same as for the analogous ketones without azido substituents (see Scheme 1).22 Thus the first excited triplet state of azides 1a-c can be expected to have an n,π*configuration whereas in azides 1d-f the π,π* excited state is lower in energy. We speculate that in azides 1, the R-cleavage comes from the n,π* triplet excited states whereas the triplet energy transfer takes place from the π,π* excited state. Photolysis of azides 1 yields a similar amount of R-cleavage products, presumably since the first and second triplet excited states are in equilibrium with each other,17 and thus we observe reactivity from both. Even though azides 1d-f have the first excited triplet state with a π,π* configuration, they do not yield less of R-cleavage products, because the triplet energy of their π,π* state is lower than that in azides 1a-c and thus the triplet energy transfer is more endothermic and slower, allowing for the R-cleavage to compete. Azides 1 can be compared to substituted valerophenone derivatives, which undergo intramolecular H-atom abstraction from their n,π* excited triplet state, whether the electron configuration of the first excited triplet state is n,π* or π,π*.17 However, it possible that triplet alkyl nitrenes 2 fragment into benzoyl and imine radicals and thus care should be taken in correlating the product ratio from azides 1 with the electron configuration of the lowest triplet excitated state of the ketone. Laser Flash Photolysis: Transient Spectra. We studied azide 1a by laser flash photolysis in methylene chloride (Lambda Physik, excimer laser, 17 ns, 150 mJ, 308 nm), recording the spectrum immediately after the laser pulse over a window of 600 ns. A moderately broad transient with a maximum around 300 nm was observed upon irradiating azide 1a (Figure 1). We have assigned this absorbance to the triplet alkyl nitrene 2a (λmax ∼ 300 nm) and benzoyl radical (λmax ) 370 nm) based on the following: The most intense absorption band near 300 nm in both triplet phenyl nitrene and NH radicals has been attributed mainly to the nz f ny transition on the nitrogen (17) (a) Wagner, P. J.; Park, B.-S. Photoinduced Hydrogen Atom Abstraction by Carbonyl Compound. In Org. Photochem. 1991, 11, 227. Ed. A. Padwa, New York, M. Dekker. (b) Wagner, P. J.; Sibert, E. J. J. Am. Chem. Soc. 1981, 103, 7329. (c) Wagner, P. J.; Kemppainen, A. E.; Schott, H. E. J. Am. Chem. Soc. 1973, 95, 5604. (18) Huggenberger, C.; Lipscher, J.; Fischer, H. J. Phys. Chem. 1980, 84, 3467. (19) Gudmundsdottir, A. D. Unpublished results. Laser flash photolysis of β-azides propiophenone gives a transient spectrum that does not have any benzoyl radical absorption present. (20) Fairchild, P. W.; Smith, G. P.; Crosley, D. R.; Jeffries, J. B. Chem. Phys. Lett. 1984, 107, 181. (21) Gritsan, N. P.; Zhu, Z.; Hadad, C. M.; Platz, M. S. J. Am. Chem. Soc. 1999, 121, 1202. (22) Murov, S. T.; Carmichael, I.; Hug, G. L. Handbook of Photochemistry, 2nd ed.;, Marcel Dekker: New York, 1993.
Photolysis of R-Azidoacetophenones
FIGURE 1. Transient spectra of azide 1a. SCHEME 7
atom (see Scheme 7) and does not involve the benzylic π-system in phenyl nitrene.20,21 Therefore, we can expect triplet alkyl nitrene to absorb somewhat similarly as triplet phenyl nitrenes despite the fact that alkyl nitrene lacks a π-system. Furthermore, the absorption spectra of triplet trifluoromethyl nitrene and triplet methyl nitrene in matrices also have absorptions around 300 nm.8,9 Triplet methyl nitrene has maximum absorption at 280 nm in matrices. The laser flash apparatus used in our studies is limited to detection above 300 nm and thus we cannot determine whether nitrene 2a also absorbs strongly at 280 nm in solution. Benzoyl radicals have a weak UV absorbance with λmax ) 370 nm that extents to 600 nm thus giving the transient spectra of 1a a broad appearance compared to that of the triplet aryl nitrenes.18,19 On a microsecond time scale, the transient with a maximum absorbance around 300 nm is unaffected in solutions saturated with oxygen. It can be concluded that the rate of interception of the triplet alkyl nitrene with oxygen must be slower than 104 s-1. In contrast, when we obtained transient spectra 200 ns after the laser pulse in oxygen-saturated solutions, the benzoyl radical absorption is depleted (see Figure 2). The transient spectrum of azide 1a in methylene chloride does not exhibit absorption due to the triplet ketone, presumably because the triplet energy is efficiently transferred to the azido group at a rate faster than the time resolution of the laser flash apparatus. In contrast (Figure 3), the transient spectrum of azide 1f, taken immediately after the laser pulse in a 200-ns window, reveals transient absorption of both the triplet ketone (λmax ) 420-450 nm) and the triplet alkyl nitrene (λmax ≈ 300 nm). A transient spectrum acquired 6900 ns after the laser pulse, in a 600-ns window, shows only the triplet alkyl nitrene and benzoyl radical absorption. Assuming that the triplet energy of the ketone moieties in azides 1 is the same as in the analogous ketones, the triplet energy of the ketone moiety in 1a and 1f is 311 and 254 kJ/mol, respectively.22 Thus the lower triplet ketone energy in 1f makes the triplet energy transfer less
FIGURE 2. Transient spectrum of 1a 200 ns after the laser pulse under O2 and Ar.
FIGURE 3. Transient spectra of 1f: (A) immediately after the laser pulse and (B) 6900 ns after the laser pulse.
FIGURE 4. Kinetic traces (A) of 1a at 320 nm and (B) of 1f at 320 and 430 nm.
exothermic and slower than that in 1a, enabling the triplet ketone absorption and the triplet nitrene to be detected. Kinetics. The absorption of nitrene 2a was monitored at 320 nm and it formed faster than the time resolution of the laser flash photolysis apparatus (see Figure 4a). Therefore energy transfer from the triplet ketone to the azido group must be faster than 17 ns. Stern-Volmer J. Org. Chem, Vol. 68, No. 21, 2003 7955
Singh et al.
FIGURE 6. Syn and gauche conformers of 1a. FIGURE 5. Rate of triplet energy transfer in azides 1.
treatment of the quenching of the absorbance at 320 nm with isoprene in ethanol gave a linear plot with a slope of 0.9. Assuming that the energy transfer is diffusion controlled or kq is between 1 and 10 × 109 M-1 s-1, the lifetime of the triplet ketone can be estimated to be between 0.9 and 0.09 ns. For azide 1f, the triplet ketone absorbance at 430 nm decays with a rate of 2 × 106 s-1 in methylene chloride (see Figure 4b) and the nitrene absorption at 320 nm forms at the same rate. The triplet ketone moieties in azides 1a and 1f are substantially shorter lived than the triplet ketone in the analogous acetophenone and p-phenylacetophenone, which have a lifetime of 0.3 and 5 µs, respectively, in methylene chloride. The short triplet lifetimes of the ketone azides are presumably due to efficient energy transfer to the azido moiety. The triplet energy transfer in azide 1f must be endothermic, which allows the energy transfer in azide 1f to be measured. We set out to estimate the triplet energy of the azido group in azides 1 by measuring the rate of triplet energy transfer in azides 1 and comparing it to their triplet ketone energies. For azides 1b and 1c we used Stern-Volmer quenching with isoprene to estimate the lifetime of the triplet ketone, whereas for azides 1d and 1e we measured directly the rate of the energy transfer. As these compounds do not show any phosphorescence emission due to the efficient intramolecular quenching of the triplet ketone, we estimate that the triplet energies of the ketones are the same as those of the analogous ketones without R-azido substituents.22 In Figure 5 we have plotted the rates of triplet energy transfer versus the triplet energy of the ketone. From this graph we can approximate that the triplet energy of the alkyl azide group must be slightly higher than 310 kJ/mol. This is in an agreement with Lewis et al., who estimated the triplet energy of simple alkyl azides to be between 314 and 335 kJ/mol.23
Theoretical Calculations To identify possible intermediates in this study, we did theoretical calculations at several different levels of theory for azide 1a and nitrene 2a. Geometry optimization of azide 1a at the B3LYP/6-311G level of theory resulted in two low-energy conformations with the azide group in either the syn or the gauche position (see Figure 6). The calculated energy difference between these two conformers is 0.4 kcal/mol. The geometries of the syn, gauche, and anti conformers of triplet nitrene 2a were also optimized at the B3LYP/6-31G* level of theory (see Figure 7). The harmonic frequency calculations indicated that the syn and gauche conformers had all positive frequencies, demonstrating that these conformers correspond to local
7956 J. Org. Chem., Vol. 68, No. 21, 2003
FIGURE 7. Syn and gauche conformers of 2a. SCHEME 8
energy minima on the potential energy surfaces. In contrast, the anti conformer had one negative frequency for a vibration associated with the rotation of the -CH2- group. This indicates that the anti conformer corresponds to either a saddle point or maximum on the potential energy surface. As such, this conformer was not investigated any further. The calculated energy difference between the gauche and the syn conformers was 1.4 kcal/mol. Triplet nitrene 2a was also investigated at the CASSCF(6,6)/6-31G* level of theory. The active space consisted of six electrons and six molecular orbitals, which were the two halffilled p orbitals on the nitrogen, the lowest occupied π and σ orbitals, and the corresponding antibonding orbitals. At this level of theory, the syn and gauche conformers again were found to be local minima on the potential energy surfaces, with an energy difference of 0.8 kcal/mol. Infrared spectra, including those of compounds with 13C (R position) and 15N labels (see Scheme 8), were calculated for nitrene 2a at both the B3LYP/ 6-31G* and CASSCF(6,6)/6-31G* levels of theory, and will be discussed below.
Matrix Isolation of Photolysis of 1a We deposited azide 1a into argon matrices in several different experiments, entraining the vapor pressure over a sample of the pure azide in flowing argon and depositing the resulting sample on a 14 K cryogenic surface. The infrared spectrum of such a sample showed a strong, broad azide band at 2111 cm-1 with a shoulder at 2104 cm-1 (see Figure 8). These frequencies fit reasonably well with the calculated frequencies for the syn and gauche conformers of 1a, respectively, and thus we assign them accordingly. On the basis of the relative intensities of the 2104- and 2111-cm-1 features, we estimated the syn conformer to be approximately three times as abundant in the argon matrix as the gauche conformer. Irradiation of azide 1a in argon matrices led to complete destruction of bands due to the azide, and the
Photolysis of R-Azidoacetophenones
FIGURE 8. IR spectra of azide 1a in argon matrices: before (red) and after irradiation (blue).
FIGURE 9. (I) IR spectra of 2a in argon matrices. (II) Expansion of regions A, B, and C. Blue spectra are of 2a with a 13C isotope in the R-position. Green spectra are of 2a with a 1:1 mixture of 15N and 14N isotopes. Red spectra are of 2a with normal isotopes. (III) Simulated spectra (B3LYP, 631G*) of syn (blue) and gauche (red) conformers of 2a.
growth of many new features attributable to photoproduct(s). Many of these occurred very near bands of the parent azide, which is expected since the acetophenone moiety should not be significantly affected by changes due to irradiation. This point is also validated by our calculations (see Figure 9). We observed several distinct new bands upon irradiation, at 536, 939, and 1201 cm-1 (see Figure 9). Of these, the band at 1201 cm-1 was the
most intense and distinct. All three bands shifted to lower energy upon 13C substitution, whereas the bands at 536 and 1201 cm-1 also shifted to lower energy with 15N substitution, as shown in Table 1. In contrast, the remainding product bands did not shift upon isotopic substitution, confirming that they are associated with vibrations of the acetophenone moiety. This observation limits the number of possibilities of the absorbing species. Candidates include the triplet nitrene, the singlet nitrene, imine 17, and the R-cleavage products. We can rule out imine 17 since we do not observe any N-H stretching mode between 3100 and 3500 cm-1, despite the fact that the vibrational mode should absorb strongly in this region. This allows us to exclude the formation of singlet alkyl nitrene in matrices as well. Ellison et al. have shown that there is no barrier separating the rearrangement of singlet methyl nitrene into an imine and thus we would expect singlet nitrene to undergo a 1,2-H shift to form imine 17 or intersystem cross to the ground-state triplet nitrene 2a.11 Furthermore, we can also rule out R-cleavage in argon matrices since we do not observe any benzoyl radical formation, which has a strong absorption at 1828 cm-1 in fluid solutions.12 We speculated, though, that if R-cleavage took place at low temperature then the methyl azide radical could expel a nitrogen molecule and rearrange into an imine radical. The argon matrices would hold the benzoyl and imine radicals in close approximation and allow them to recombine to form imine 16 (see Scheme 9). However, imine 16 is not formed in argon matrices, because our calculations show that the frequency of the carbonyl stretch would be shifted 27 and 18 cm-1 upon 13C and 15N substitution, respectively. We calculated the infrared spectrum of the syn and the gauche conformers of triplet nitrene 2a at several levels of theory, for comparison to these three bands. The C-N stretch and the C-C-N bending modes are expected to be very sensitive to 13C and 15N substitution, whereas the CH2 wag should be sensitive to 13C but not 15N substitution. The C-N stretches of nitrenes have been observed between 1000 and 1075 cm-1, for example, the C-N stretch of triplet methyl nitrene has been reported at 1040 cm-1 whereas the C-N stretch of tetrafluoropyridine-4-nitrene has been observed at 1272 J. Org. Chem, Vol. 68, No. 21, 2003 7957
Singh et al. TABLE 1. Comparison of Experimental and Theoretical Band Positions and Isotopic Shift for Nitrene 2a 13C
band position
15N
shift
shift
B3LYPa
CASSCF
exptl
B3LYP
CASSCF
exptl
B3LYP
CASSCF
exptl
663 909 1067
360 1002 1145
539 939 1201
-5 -7 -17
-2 -8 -16
-2 -6 -9
-1 -1 -15
-5 -1 -15
-3 0 -6
a The calculated IR frequencies are not scaled. In the Supporting Information we have listed the calculated IR frequencies for B3LYP/ 6311++G*, which are generally not scaled. There is not a significant shift in the IR frequencies of these three bands with use of either one of these basis sets.
SCHEME 9
cm-1.6,24 Our calculations place this mode between 1067 (B3LYP/6-31G*) and 1145 cm-1 (CASSCF/6-31G*). On the basis of the location of the 1201-cm-1 band and its sensitivity to both 13C and 15N substitution, we assign it to the C-N stretch of the syn conformer of nitrene 2a. Since azide 1a is primarily trapped in the syn conformation we expect it to produce primarily the syn conformer of nitrene 2a. We assigned the 939-cm-1 band to the CH2 wagging motion of the -CH2N group; the calculated position and isotopic shifts are in good agreement with the experimental values. Less is known experimentally about the C-C-N bending mode, although it should occur at substantially lower frequency. For comparison, the C-CdO bending mode of acetone occurs near 535 cm-1.25 The bending mode is calculated at 663 cm-1 at the B3LYP/6-31G* level of theory. While this is higher than the 536-cm-1 experimental band, the calculated isotopic shifts are reasonable. On the other hand, calculations at the CASSCF/6-31G* level substantially underestimated this band position, but again the calculated isotopic shifts are in good agreement. Overall, while theoretical calculations have some difficulty calculating absolute band positions for the nitrene moiety, the calculated isotope shifts for these three modes are quite good, and confirm identification of nitrene 2a. The UV/vis spectrum obtained after photolysis of azide 1a in argon matrices shows no absorption due to benzoyl radical formation (see Figure 10) and it does resemble the UV absorption spectrum of triplet methyl nitrene.6 When comparing the transient spectra of nitrene 2a obtained with the laser flash apparatus in fluid solutions and in matrices it must be kept in mind that the laser flash apparatus is limited to detection above 300 nm. Thus the laser flash photolysis catches the trail of the UV absorption of nitrene 2a, whereas the transient spectrum of 2a in argon matrices has λmax at 280 nm.
benzoyl radicals is favored whereas at lower temperature more of the triplet alkyl nitrenes are formed. The triplet alkyl nitrenes 2 are intercepted with both molecular oxygen and benzoyl radicals. The triplet alkyl nitrenes can be detected directly with laser flash photolysis and they have a lifetime longer than microseconds in solution. The triplet alkyl nitrene intermediates were further characterized by obtaining their UV and IR spectrum in argon matrices. The UV absorption spectrum of triplet alkyl nitrene 2a in argon matrices has maximum absorption at 280 nm. The IR spectrum of nitrene 2a in argon matrices has a weak C-N stretch at 1201 cm-1, which shifted on 15N and 13C isotope labeling.
Conclusion
Experimental Section
We have shown that irradiation of azides 1 leads to the formation of benzoyl radicals and triplet alkyl nitrenes 2. At higher temperatures R-cleavage to form
Calculations.: The geometries of azide 1a were optimized by the B3LYP method as implemented in the Gaussian 98 programs, using the 6-311G basis set. The geometries and IR spectra of nitrene 2a were optimized by the B3LYP and CASSCF theory, using the 6-31G* basis set. The calculated IR frequencies are not scaled. All the calculations were preformed with the Gaussian 98 software package.32
(23) Lewis, F. D.; Saunders, W. H., Jr. J. Am. Chem. Soc. 1968, 90, 7033. (24) Chapyshev, S. V.; Kuhn, A.; Wong, M. W.; Wentrup, C. J. Am. Chem. Soc. 2000, 122, 1572. (25) Marcou, A.; Roche, D. C. R. Acad. Sci. Ser. B 1973, 276, 599.
7958 J. Org. Chem., Vol. 68, No. 21, 2003
FIGURE 10. The UV absorbance spectra of azide 1a in argon matrices: (a) before photolysis, (b) after photolysis, and (c) the difference of spectrum b - spectrum a.
(26) Ault, B. S. J. Am. Chem. Soc. 1978, 100, 2426.
Photolysis of R-Azidoacetophenones Laser Flash Photolysis. Laser flash photolysis was carried out with an Excimer laser (308 nm, 17 ns). The system has been described in detail elsewhere.27 The rate of the energy transfer for azides 1d-f was measured directly. We did not detect the absorption of the triplet excited state of the ketone chromophore in azides 1a-c. However, we succefully manged to reduce the nitrene absorption at 320 nm with isoprene, because the precursor or the triplet excited ketone is being quenched. Standard Stern-Volmer treatment of the quenching of the absorbance at 320 nm with isoprene gave a linear plot. By assuming that the energy transfer from the triplet excited state of the ketone to isoprene was diffusion controlled, or kq is 5 × 109 M-1 s-1, the lifetime of the triplet ketone could be estimated from the slope of the Stern-Volmer graph.28 Matrix Isolations. Matrix isolation studies were done on conventional equipment.26 General Procedures. Photolysis were carried out with a 450-W mercury lamp through Pyrex filters. Toluene was dried and distilled over sodium prior to use. Preparation of Azides 1a-f. Azides 1 were synthesized as described.28 In a typical reaction 2-bromoacetophenone (2.0 g, 0.010 mol) was dissolved in ethanol (50 mL) and glacial acetic acid (2 mL). To this mixture was added a solution of sodium azide (1.30 g, 0.204 mol) in water (5 mL) and the resulting mixture was refrigerated overnight. The reaction was neutralized with aqueous saturated sodium bicarbonate solution and extracted with ethyl acetate. The organic layer was washed with brine and dried over MgSO4 and the solvent was removed under vacuum. The resulting solid was recrystallized from ethanol in an ice bath (1.32 g, 8.2 mmol, 82% yield). The 1H NMR, IR spectra and the melting point of 2-azido-1phenylethanone, 1a, are identical with those in the literature.10,28 The labeled 1aC was synthesized in the same manner from 2-bromo-1-13C-acetophenone. The 1:1 mixture of 1aN1:1aN3 was prepared by the same method with 15N-labeled sodium azide. Azides 1b-f were synthesized in approximately 80% yield by the same procedure as azide 1a. The 1H NMR, IR spectra, and melting points of 2-azido-(4′-bromophenyl)ethanone, 1b,33 2-azido-(4′-chlorophenyl)ethanone, 1c,10,28 and 2-azido-1-benzo[1,3]dioxol-5-ylethanone, 1e,34 are identical with those in the literature. 2-Azido-(4′-cyanophenyl)ethanone, 1d. Mp (ethyl acetate/ hexanes) 127-130 °C. IR (KBr) 2232, 2107, 1693 cm-1. 1H NMR (250 MHz, CDCl3) δ 4.57 (s, 2 H), 7.82 (d, 8 Hz, 2H), (27) Gritsan, N. P.; Zhai, H. B.; Yuzawa, T.; Karweik, D.; Brooke, J.; Platz, M. S. J. Phys. Chem. A 1997, 101, 2833. (28) Turro, N. J. Modern Molecular Photochemistry; University Science Books: Sausalito, CA, 1991; Chapter 8. (29) (a) Boyer, J. H.; Straw, D. J. Am. Chem. Soc. 1952, 74, 4506. (b) Boyer, J. H.; Straw, D. J. Am. Chem. Soc. 1953, 75, 1642. (30) (a) Hegarty, D.; Robb, M. A. Mol. Phys. 1979, 38, 1795. (b) Robb, M. A.; Niazi, U. Rep. Mol. Theory 1990, 1, 23. (c) Anderson, K.; Roos, B. O. In Modern Electron Structure Theory; World Scientific Publishing: Singapore, 1995; Part 1, Vol. 2, p 55. (31) (a) Gross, E. K. U.; Kohn, W. Adv. Quantum Chem. 1990, 21, 255. (b) Casida, M. E. In Recent Advances in Density Functional Methods; World Scientific Publishing: Singapore, 1995; Vol. 1, 55. (c) Wiberg, K. B.; Stratmann, R. R.; Frisch, M. J. Chem. Phys. Lett. 1998, 297, 60. (32) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Cliford, S.; Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick, D. K.; Rbuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.; Stafanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Goperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Andres, J. L.; Head-Gordon, M.; Replogle, E. S.; Pople, J. A. Gaussian 94, Revision D.2; Gaussian, Inc.: Pittsburgh, PA, 1998. (33) Nair, V.; Nair, L. G.; George, T. G.; Augustine, A. Tetrahedron 2000, 56, 7607. (34) Ackrell, J.; Muchowski, J. M. J. Org. Chem. 1986, 51, 3374.
7.99 (d, 8 Hz, 2H) ppm. 13C NMR (60 MHz, CDCl3) δ 55.2, 117.5, 128.5, 132.5, 137.3, 192.2 ppm. MS (ESI) 159.1 (M+ N2). 2-Azido-(4′-biphenyl-4-yl)ethanone, 1f. Mp (ethyl acetate/ hexanes) 78-82 °C (lit.28 mp 88-88.5 °C). IR (KBr) 2099, 1683 cm-1. 1H NMR (250 MHz, CDCl3) δ 4.59 (s, 2H), 7.45 (m, 3H), 7.64 (d, 5 Hz, 2H), 7.71 (d, 8 Hz, 2H), 7.98 (d, 8 Hz, 2H) ppm. 13 C NMR (60 MHz, CDCl3) δ 54.59, 126.7, 126.9, 127.6, 128.1, 128.6, 132.6, 138.9, 146.2, 192.8 ppm. Preparative Photolysis of 1. Formation of acetophenone 4 and benzaldehyde 5 was confirmed by authentic injection on GC and GC-MS and HPLC. Products 3a-f were isolated and characterized. Photolysis of 1a. A solution of azide 1a (80 mg, 0.5 mmol) in toluene (2 mL) was degassed with argon and irradiated for 26 h. GC analysis of the reaction mixture showed no remaining 1a. The solvent was removed under vacuum and the resulting oil purified on a silica column eluted with 10% ethyl acetate in hexane. The column yielded three fractions: benzaldehyde (15 mg, 0.14 mmol, 28%), acetophenone (11 mg, 0.09 mmol, 18%), and amide 3a (26 mg, 0.11 mmol, 44%). The total isolated yield of photoproducts is 90%. The 1H NMR, IR, and melting point of N-(2-oxo-2-phenyl-ethyl)benzamide, 3a, were identical with those in the literature.10 Photolysis of 1b. A solution of 1b (916 mg, 1.5 mmol) in toluene (80 mL) was photolyzed overnight. GC analysis of the reaction mixture showed no remaining starting material. The solvent was removed under vacuum and the resulting oil was purified on a silica column eluted with hexane in ethyl acetate. The major product isolated was 4-chloro-N-[2-(4-chlorophenyl)2-oxoethyl]benzamide, 3b (65 mg, 0.21 mmol, 30% yield). Mp (ethyl acetate/hexane) 172-174 °C. IR (KBr) 3367, 1691, 1635, 1094 cm-1. 1H NMR (250 MHz, CDCl3) δ 4.92 (d, 6 Hz, 2H), 7.23 (s, 1H), 7.47 (m, 6 Hz, 5H), 7.80 (d, 8 Hz, 2H), 8.0 (d, 8 Hz, 2H) ppm.13C NMR (60 MHz, CDCl3) δ 47.0, 128.8, 129.1, 129.1, 129.2, 129.6, 129.7, 132.5, 133.0, 138.4, 141.1, 166.5, 193.3 ppm. GCMS 307/309 (M+, 2%), 168/167 (5%), 139/141 (100%). Photolysis of 1c. A solution of 1c (200 mg, 0.84 mmol) was photolyzed until all the GC analyzis showed only 5% remaining starting material. The solvent was removed under vacuum and the resulting oil purified with silica column eluted with 5% ethanol in hexane. Three fractions were collected: 3c (77 mg, 0.20 mmol, 47% yield), 4c (12 mg, 0.06 mmol, 7% yields), and 5c (3 mg, 0.017 mmol, 2% yields). The 1H NMR, 13C NMR, IR, and melting point of 4-bromo-N-[2-(4-bromophenyl)-2-oxoethyl]benzamide, 3c, were identical with those in the literature.10 Photolysis of 1d. A solution of 1d (900 mg, 4.8 mmol) in toluene (20 mL) was bubbled with argon and irradiated overnight. Insoluble polymers formed upon photolysis which were filtered and discarded. HPLC showed formation of 3d and 11d. The solvent was removed under vacuum to yield oil that was purified on silica gel eluted with ethyl acetate in hexane. 4-Cyano-N-[2-(4-cyanophenyl)-2-oxyethyl]benzamide,, 3d, was isolated in 6% yield (39.2 mg, 0.14 mmol). IR (CHCl3) 3407, 2230, 1704, 1641, 1607, 1404 cm-1. 1H NMR (250 MHz, CDCl3) δ 8.14 (d, 8 Hz, 2H), 8.05 (d, 8 Hz, 2H), 7.9-7.6 (m, 4 H), 7.25 (s, 1H), 4.98 (d, 3 Hz, 2H) ppm. 13C NMR (60 MHz, DMSO-d6) δ 47.0, 118.2, 118.4, 128.2, 128.3, 132.6, 132.7, 138.2, 139.7, 165.5, 195.0 ppm. HRMS (ESI) calcd for C17H10N3O2 (M+ - 1) 288.0773, found 288.0773. 4-(2-Benzylamino-acetyl)benzonitrile, 11d, was isolated in 1% yield (13.5 mg, 0.054 mmol). IR (CHCl3) 3406, 2233, 1667, 1607, 1521, 1422 cm-1. 1H NMR (250 MHz, CDCl3) δ 4.65 (s, 2H), 4.84 (s, 2H) ppm. 13C NMR (60 MHz, CDCl3) δ 38.7, 56.8, 115.6, 117.4, 127.7, 127.8, 128.7, 132.7, 132.8, 134.5, 137.7, 197.1 ppm. Photolysis of 1e. A solution of 3e (250 mg, 1.22 mmol) in toluene (50 mL) was photolyzed overnight. The solvent was removed under vacuum and the resulting oil purified on a silica column to yield four fractions: 1e (40 mg, 0.20 mmol,
J. Org. Chem, Vol. 68, No. 21, 2003 7959
Singh et al. 16%), 4e (15 mg, 0.091 mmol, 7% yield), 5e (35 mg, 0.23 mmol, 19% yield), and 3e (60 mg, 0.18 mmol, 30% yield). The physical properties of (2-benzo[1,3]dioxol-5-yl-2-oxo-ethyl)amide benzo[1,3]dioxole-5-carboxylic acid, 3e, are as follows. IR (CHCl3) 1677, 1657, 1604, 1254 cm-1. 1H NMR (250 MHz, CDCl3) δ 4.84 (d, 4 Hz, 2H), 6.04 (s, 2H), 6.08 (s, 2H), 6.88 (m, 2H), 7.15 (s, 1H), 7.40 (m, 4H), 7.63 (d, 8 Hz, 2H) ppm. 13C NMR (60 MHz, CDCl3) δ 192.3, 166.5, 152.7, 150.5, 148.5, 148.0, 129.2, 128.2, 124.4, 121.8, 108.3, 108.0, 107.9, 107.6, 102.1, 101.7, 46.6 ppm. HRMS calcd for C17H13NO6Na, [M + 23]+, 350.0641, found 350.0665. Photolysis of 1f. A solution of 1f (298 mg, 1.25 mmol) in toluene (60 mL) was photolyzed overnight. A TLC showed no remaining starting material. The solvent was removed under vacuum and the resulting oil purified on a silica column eluted with ethyl acetate in hexane. Three fractions were collected: 4f (2.4 mg, 0.013 mmol, 1%), p-phenyl acetophenone 5f (35 mg, 0.18 mmol, 14% yields), and 3f (45 mg, 0.12 mmol, 20% yields). The spectroscopic characterization of biphenyl-4carboxylic acid (2-biphenyl-4-yl-2-oxoethyl)amide, 3f, is listed below. Mp 257-8 °C dec (lit.34 mp 259 °C). IR (KBr) 3388, 1646, 1603 cm-1. 1H NMR (250 MHz, CDCl3) δ 4.81 (d, 4 Hz, 2H), 7.39 (m, 10H), 7.72 (d, 8 Hz, 3H), 7.87 (d, 8 Hz, 3H) ppm. Analytical Photolysis of Azides 1. Azides 1 were dissolved in toluene, degassed by bubbling argon through the solution, and photolyzed through Pyrex filter. The conversion of the starting material was kept below 20%. HPLC, GC analysis, and 1H NMR of the reaction mixtures were used to estimate the product ratio for 3, 4, and 5. Photolysis of 1-Azidoadmantane and Methyl Benzoin, 8. A solution of 8 (17 mg, 0.1 mmol) and 1-azidoadmantane (22.6 mg, 0.1 mmol) in 2 mL of toluene was degassed with argon and photolyzed with a 450-W Mercury arc lamp through a 349-nm filter. A solution of 8 (0.1 M) in toluene was also prepared and photolyzed simultaneously. The major products coming from photolysis of 8 were characterized by GC/MS as 5a, benzil (9), 1,2-diphenylethanone (10), 1-methoxy-1,2diphenylethane (12), and 1,2-dimethoxy-1,2-diphenylethane (11). Photolysis of 1a in Toluene at Low Temperature. A solution of 1a (0.02 M) and hexadecane (0.005 M) in toluene
7960 J. Org. Chem., Vol. 68, No. 21, 2003
was photolyzed at -63 and 100 °C. The reaction mixture was analyzed by GC and GC/MS. Formation of 1,2-diphenylethanone (9) and 1,4-diphenylbutane-1,4-dione (12) was verified with commercially available authentic samples. Formation of N-benzyl-N-(2-oxo-2-phenylethyl)benzamide (13) and 2-benzylamino-1-phenylethanone (11) was established by synthesizing these compounds following literature procedures.36 Photolysis of 1a in Toluene Saturated with Oxygen. An oxygen-saturated solution of 1a (0.01 M) in toluene was photolyzed at 30 and -63 °C. GC analysis of the reaction mixture showed formation of benzoic acid, acetophenone, 2-nitrophenylethanone, and 3a along with unreacted starting material. The formation of benzoic acid, acetophenone (4a), 2-nitrophenylethanone (15), and 3a was verified by GC/MS and authentic samples.36
Acknowledgment. A.D.G. is grateful to the University of Cincinnati Research Council Faculty Research Support. We also acknowledge the finical support from the National Science Foundation (CHE No. 0093622) and the Petroleum Research Foundation (ACS-PRF No. #35809). S.M.M. thanks the University of Cincinnati Research Council for financial support. We thank Professor M. S. Platz at The Ohio State University for the use of his laser flash apparatus. Supporting Information Available: Cartesian coordinates, number of imaginary frequencies, IR frequencies, and total energy of 1a, 2a, and 16; the Stern Volmer graph for isoprene quenching of azides 1a-c and 1H NMR spectra of products 3; the rate of intramolecular energy transfer in 1. This material is available free of charge via the Internet at http://pubs.acs.org. JO034674E (35) Balaban, A. T.; Bally, I.; Frangopol, P. T.; Bacescu, M.; Cioranescu, E.; Birladeanu, L. Tetrahedron 1963, 19, 169. (36) Fuhrman, H.; Haber, H.; Haupt, M.; Henning, H.-G. J. Prakt. Chem. 1982, 6, 1055.
